Hi Andrey. > > > + > > > + chip->miibus.read = mv88e6xxx_mdio_read; > > > + chip->miibus.write = mv88e6xxx_mdio_write; > > > > The function pointers are hardcoded here. > > But we have them in chip->info->ops - where we can > > have chip specific variants. > > I assume it would be more correct to copy them from the ops structure? > > > > I am not sure I see why that would be. Mv88e6xxx_mdio_read() and > mv88e6xxx_mdio_write() are both wrappers around > chip->info->ops->phy_read/phy_write that also have code to handle the > case when either phy_read/phy_write are not specified. I should stop reading patches late. I read the above as function pointers to the functions used to read the phy, and not the general mdio_read/write functions. So as you points outs this is fine. Sorry for the noise. I also browsed through the other patches in this set, and everything else looked good. But then some parts I was not familiar with. Sam _______________________________________________ barebox mailing list barebox@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/barebox
